The present invention relates to the fields of protein engineering, plant molecular biology and pest control. More particularly the invention relates to novel engineered hybrid proteins having insecticidal activity, nucleic acids whose expression results in the insecticidal proteins, and methods of making and methods of using the insecticidal proteins and corresponding nucleic acids to control insects.
Insect pests are a major cause of crop losses. In the US alone, billions of dollars are lost every year due to infestation by various genera of insects. In addition to losses in field crops, insect pests are also a burden to vegetable and fruit growers, to producers of ornamental flowers, and they are a nuisance to gardeners and homeowners.
Species of corn rootworm are considered to be the most destructive corn pests. In the United States, the three important species are Diabrotica virgifera virgifera, the western corn rootworm, D. longicornis barberi, the northern corn rootworm and D. undecimpunctata howardi, the southern corn rootworm. Only western and northern corn rootworms are considered primary pests of corn in the US Corn Belt. An important corn rootworm pest in the Southern US is the Mexican corn rootworm, Diabrotica virgifera zeae. Corn rootworm larvae cause the most substantial plant damage by feeding almost exclusively on corn roots. This injury has been shown to increase plant lodging, to reduce grain yield and vegetative yield as well as alter the nutrient content of the grain. Larval feeding also causes indirect effects on corn by opening avenues through the roots for bacterial and fungal infections which lead to root and stalk rot diseases. Adult corn rootworms are active in cornfields in late summer where they feed on ears, silks and pollen, thus interfering with normal pollination.
Corn rootworms are mainly controlled by intensive applications of chemical pesticides, which are active through inhibition of insect growth, prevention of insect feeding or reproduction, or cause death. Good corn rootworm control can thus be reached, but these chemicals can sometimes also affect other, beneficial organisms. Another problem resulting from the wide use of chemical pesticides is the appearance of resistant insect varieties. Yet another problem is due to the fact that corn rootworm larvae feed underground thus making it difficult to apply rescue treatments of insecticides. Therefore, most insecticide applications are made prophylactically at the time of planting. This practice results in a large environmental burden. This has been partially alleviated by various farm management practices, but there is an increasing need for alternative pest control mechanisms.
Biological pest control agents, such as Bacillus thuringiensis (Bt) strains expressing pesticidal toxins like δ-endotoxins (delta-endotoxins; also called crystal toxins or Cry proteins), have also been applied to crop plants with satisfactory results against primarily lepidopteran insect pests. The δ-endotoxins are proteins held within a crystalline matrix that are known to possess insecticidal activity when ingested by certain insects. The various δ-endotoxins have been classified based upon their spectrum of activity and sequence homology. Prior to 1990, the major classes were defined by their spectrum of activity with the Cry1 proteins active against Lepidoptera (moths and butterflies), Cry2 proteins active against both Lepidoptera and Diptera (flies and mosquitoes), Cry3 proteins active against Coleoptera (beetles) and Cry4 proteins active against Diptera (Hofte & Whitely, 1989, Microbiol. Rev. 53:242-255). A new nomenclature was developed in 1998 which systematically classifies the Cry proteins based on amino acid sequence homology rather than insect target specificities (Crickmore et al. 1998, Microbiol. Molec. Biol. Rev. 62:807-813).
The spectrum of insecticidal activity of an individual δ-endotoxin from Bt is quite narrow, with a given δ-endotoxin being active against only a few species within an Order. For instance, a Cry3A toxin is known to be very toxic to the Colorado potato beetle, Leptinotarsa decemlineata, but has very little or no toxicity to related beetles in the genus Diabrotica (Johnson et al., 1993, J. Econ. Entomol. 86:330-333). According to Slaney et al. (1992, Insect Biochem. Molec. Biol. 22:9-18) a Cry3A toxin is at least 2000 times less toxic to southern corn rootworm larvae than to the Colorado potato beetle. It is also known that Cry3A has little or no toxicity to the western corn rootworm or northern corn rootworm.
Specificity of the δ-endotoxins is the result of the efficiency of the various steps involved in producing an active toxic protein and its subsequent interaction with the epithelial cells in an insect mid-gut. To be insecticidal, most known δ-endotoxins must first be ingested by the insect and proteolytically activated to form an active toxin. Activation of the insecticidal crystal (Cry) proteins is a multi-step process. After ingestion, the crystals must first be solubilized in the insect gut. Once solubilized, the δ-endotoxins are activated by specific proteolytic cleavages. The proteases in the insect gut can play a role in specificity by determining where the δ-endotoxin is processed. Once the δ-endotoxin has been solubilized and processed it binds to specific receptors on the surface of the insects' mid-gut epithelium and subsequently integrates into the lipid bilayer of the brush border membrane. Ion channels then form disrupting the normal function of the midgut eventually leading to the death of the insect.
In Lepidoptera, which have alkaline pH guts, gut proteases process δ-endotoxins for example, Cry1Aa, Cry1Ab, Cry1Ac, Cry1B and Cry1F, from 130-140 kDa protoxins to toxic proteins of approximately 60-70 kDa. Processing of the protoxin to toxin has been reported to proceed by removal of both N- and C-terminal amino acids with the exact location of processing being dependent on the specific δ-endotoxin and the specific insect gut fluids involved (Ogiwara et al., 1992, J. Invert. Pathol. 60:121-126). Thus activation requires that the entire C-terminal protoxin tail region be cleaved off. This proteolytic activation of a δ-endotoxin can play a significant role in determining its specificity.
Coleopteran insects have guts that are more neutral to acidic and coleopteran-specific δ-endotoxins are similar to the size of the activated lepidopteran-specific toxins. Therefore, the processing of coleopteran-specific δ-endotoxins was formerly considered unnecessary for toxicity. However, data suggests that coleopteran-active δ-endotoxins are solubilized and proteolyzed to smaller toxic polypeptides. A 73 kDa Cry3A δ-endotoxin protein produced by B. thuringiensis var. tenebrionis is readily processed in the bacterium at the N-terminus, losing 49-57 residues during or after crystal formation to produce the commonly isolated 67 kDa form (Carroll et al., 1989, Biochem. J. 261:99-105). McPherson et al., (1988, Biotechnology 6:61-66) also demonstrated that a native cry3A coding sequence contains two functional translational initiation codons in the same reading frame, one coding for a 73 kDa protein and the other coding for a 67 kDa protein starting at Met-1 and Met-48 respectively, of the deduced amino acid sequence. Both proteins then can be considered naturally occurring full-length Cry3A proteins.
As more knowledge has been gained as to how the δ-endotoxins function, attempts to engineer δ-endotoxins to have new activities have increased. Engineering 8-endotoxins was made more possible by solving the three dimensional structure of Cry3A in 1991 (Li et al., 1991, Nature 353:815-821). Li et al. determined that a Cry3A protein has three structural domains: the N-terminal domain I, from residues 58-290, consists of 7 α-helices, domain II, from residues 291-500, contains three β-sheets in a so-called Greek key-conformation, and the C-terminal domain III, from residues 501-644, is a β-sandwich in a so-called jellyroll conformation. The three dimensional structure for the lepidopteran active Cry1Aa toxin has also been solved (Grochulski et al., 1995, J. Mol. Biol. 254:447-464). The Cry1Aa toxin also has three domains: the N-terminal domain I, from residues 33-253, domain II from residues 265-461, and domain III from residues 463-609 with an additional outer strand in one of the β-sheets from by residues 254-264. If the Cry3A and Cry1Aa structures are projected on other Cry1 sequences, domain I runs from about amino acid residue 28 to 260, domain II from about 260 to 460 and domain III from about 460 to 600. See, Nakamura et al., Agric. Biol. Chem. 54(3): 715-724 (1990); Li et al., Nature 353: 815-821 (1991); Ge et al., J. Biol. Chem. 266(27): 17954-17958 (1991); and Honee et al., Mol. Microbiol. 5(11): 2799-2806 (1991); each of which are incorporated herein by reference. Thus, it is now known that based on amino acid sequence homology, known Bt δ-endotoxins have a similar three-dimensional structure comprising three domains.
The toxin portions of Bt Cry proteins are also characterized by having five conserved blocks across their amino acid sequence numbered CB1 to CB5 from N-terminus to C-terminus, respectively (Hofte & Whiteley, supra). Conserved block 1 (CB1) comprises approximately 29 amino acids, conserved block 2 (CB2) comprises approximately 67 amino acids, conserved block 3 (CB3) comprises approximately 48 amino acids, conserved block 4 (CB4) comprises approximately 10 amino acids and conserved block 5 (CB5) comprises approximately 12 amino acids. The sequences before and after these five conserved blocks are highly variable and thus are designated the “variable regions,” V1-V6. Domain I of a Bt δ-endotoxin typically comprises variable region 1, conserved block 1, variable region 2, and the N-terminal 52 amino acids of conserved block 2. Domain II typically comprises approximately the C-terminal 15 amino acids of conserved block 2, variable region 3, and approximately the N-terminal 10 amino acids of conserved block 3. Domain III typically comprises approximately the C-terminal 38 amino acids of conserved block 3, variable region 4, conserved block 4, variable region 5, and conserved block 5. The Cry1 lepidopteran active toxins, among other delta-endotoxins, have a variable region 6 with approximately 1-3 amino acids lying within domain III.
Many Bt strains and δ-endotoxins are active against different insect species and nematodes. However, relatively few of these strains and toxins have activity against coleopteran insects. Further, most of the now known native coleopteran-active δ-endotoxins, for example Cry3A, Cry3B, Cry3C, Cry7A, Cry8A, Cry8B, and Cry8C, have insufficient oral toxicity against corn rootworm to provide adequate field control if delivered, for example, through microbes or transgenic plants. Therefore, other approaches for producing novel toxins active against corn rootworm need to be explored.
Lepidopteran-active δ-endotoxins have been engineered in attempts to improve specific activity or to broaden the spectrum of insecticidal activity. For example, the silk moth (Bombyx mori) specificity domain from a Cry1Aa protein was moved to a Cry1Ac protein, thus imparting a new insecticidal activity to the resulting hybrid Bt protein (Ge et al. 1989, PNAS 86: 4037-4041). Also, Bosch et al. 1998 (U.S. Pat. No. 5,736,131, incorporated herein by reference) describes Bacillus thuringiensis hybrid toxins comprising at their C-terminus domain III of a first Cry protein and at its N-terminus domains I and II of a second Cry protein. Such hybrid toxins were shown to have altered insecticidal specificity against lepidopteran insects. For example, the H04 hybrid toxin, which is also described in De Maagd et al., Appl. Environ. Microbiol. 62(5): 1537-1543 (1996), comprises at its N-terminus domains I and II of a Cry1 Ab and at its C-terminus domain III of a Cry1C. H04 is reportedly highly toxic to the lepidopteran insect Spodoptera exigua (beet armyworm) compared with the parental Cry1Ab toxin and significantly more toxic than the Cry1C parental toxin. It has also been shown that substitution of domain III of toxins, which are not active against the beet armyworm such as Cry1E and Cry1Ab, by domain III of Cry1C, which is active against beet armyworm, can produce hybrid toxins that are active against this insect. All of the hybrids disclosed in Bosch et al. use domains from lepidopteran active Cry proteins to make new toxins with lepidopteran activity. The results do suggest that domain III of Cry1C is an important determinant of specificity for beet armyworm. See also, Bosch et al., FEMS Microbiology Letters 118: 129-134 (1994); Bosch et al., Bio/Technology 12: 915-918 (1994); De Maagd et al., Appl. Environ. Microbiol. 62(8): 2753-2757 (1996); and De Maagd et al., Mol. Microbiol. 31(2): 463-471 (1999); each of which is incorporated herein by reference.
Several attempts at engineering the coleopteran-active δ-endotoxins have been reported. Chen and Stacy (U.S. Pat. No. 7,030,295, herein incorporated by reference) successfully created a corn rootworm active toxin by inserting a non-naturally occurring protease recognition site in domain I, domain III, or both domains I and III of a Cry3A protein. One of the resulting modified Cry3A proteins, designated Cry3A055, having a protease recognition site inserted in domain I, was active against several species of Diabrotica. Van Rie et al., 1997, (U.S. Pat. No. 5,659,123) engineered Cry3A by randomly replacing amino acids, thought to be important in solvent accessibility, in domain II with the amino acid alanine. Several of these random replacements confined to domain II were reportedly involved in increased western corn rootworm toxicity. However, others have shown that some alanine replacements in domain II of Cry3A result in disruption of receptor binding or structural instability (Wu and Dean, 1996, J. Mol. Biol. 255: 628-640). English et al., 1999, (Intl. Pat. Appl. Publ. No. WO 99/31248) reported amino acid substitutions in Cry3Bb that caused increases in toxicity to southern and western corn rootworm. However, of the 35 reported Cry3Bb mutants, only three, with mutations primarily in domain II and the domain I-domain II interface, were active against western corn rootworm. Further, the variation in toxicity of wild-type Cry3Bb against western corn rootworm in the same assays appear to be greater than any of the differences between the mutated Cry3Bb toxins and the wild-type Cry3Bb. Shadenkov et al. (1993, Mol. Biol. 27:586-591), made a hybrid protein by fusing amino acids 48-565 of a Cry3A protein to amino acids 526-725 of a Cry1Aa protein. Therefore, the cross-over between Cry3A and Cry1Aa sequence was in conserved block 4 located in domain III. Cry3A is very active against the Colorado potato beetle (Leptinotarsa decemlineata). However, the hybrid protein disclosed by Shadenkov et al. was not active against Colorado potato beetle even though more than 75% of the hybrid protein was made up of Cry3A sequence. Thus, the addition of only 25% of Cry1Aa sequence destroyed activity against a coleopteran insect that the parent Cry3A was active against. This suggests that hybrid proteins made by fusing portions of a coleopteran-active Cry protein, e.g. Cry3A, and a lepidopteran-active Cry protein, e.g. Cry1A, would not have activity against coleopteran insects, particularly a coleopteran insect that is not naturally susceptible to Cry3A like corn rootworm.
In view of the above discussion, there remains a need to design new and effective pest control agents that provide an economic benefit to farmers and that are environmentally acceptable. Particularly needed are proteins that are toxic to Diabrotica species, a major pest of corn, that have a different mode of action than existing insect control products as a way to mitigate the development of resistance. Furthermore, delivery of control agents through products that minimize the burden on the environment, as through transgenic plants, are desirable.